Laserfiche WebLink
<br />1" ) ~ ) <br />U ,J '.1) _...1 <br /> <br />19 <br /> <br />voir, A gross outflow rate of 50 cfs is shown superimposed on the mass in- <br />flow curve (fig. 3) at two periods of low inflow. If a proposed reservoir at the <br />site were full at the start of each of these periods and a constant gross outflow <br />of 50 cfs were maintained, the accumulated outflow would exceed the accumu- <br />lated inflow until the. two lines meet, Then the reservoir would be full again <br />if evaporation and seepage losses are neglected. Maximum withdrawal from <br />the reservoir would be reached when the inflow and outflow lines are most dis- <br />tant from each other. The greatest ordinate interval obtained by superimpos- <br />ing a 50 ds draft rate at various periods of low flow, (fig, 3), was found to be <br />the one in March 1957. This interval represents the greatest storage require- <br />ment during 1952- 57 for a continuous draft of 50 cfs. The storage requirement, <br />as computed on figure 3, equals 21,600 ds-days or 42,800 acre-feet, Dividing <br />these values by the drainage area of 922 square miles at this site shows that <br />a gross outflow of 0,054 crs per square mile would have required a maximum <br />storage capacity of 46.4 acre-feet per square mile, This relation is plotted <br />by a circle on figure 5 (subsequently presented). By constructing lines for <br />other selected outflow rates on figure 3 such as that shown for 160 cfs, addi- <br />tional storage requirements have been computed as shown by other circled <br />relations on figure 5, The circled values roughly define a relation curve of <br />storage that would have been required to maintain various rates of gross out- <br />flow during the drought of the 1950's. Similar data are shown for each gaging <br />station in operation during the period October 1951 to March 1957 (fig. 6-118 <br />subsequently presented). <br /> <br />Storage Requirements by Frequency Methods <br /> <br />The storage requirement for a specific period of record at a specific site <br />does not show how frequently that storage may be expected to prove inadequate, <br />In 1951 Hazen developed average, maximum, and minimum storage curves for <br />a group of streams and indicated that curves on a frequency basis would be <br />desirable. Smallwood and others (1954) computed annual storage requirements <br />from mass curves of actual records and arrayed the annual requirements in <br />order of magnitude to determine the percent of years when storage would be <br />adequate starting with a full reservoir each year. Hudson and Roberts (1955) <br />combined the low- flow frequency curves in southern Illinois to develop the fre- <br />quency wherein certain rates of flow could be sustained by various storage ca- <br />pacities. The U, S. Corps of Engineers (1955, p. 10) suggested development <br />of a mass curve from frequency data. Martin and Hulme (1957, p. 55) utilized <br />low-flow frequency curves to compute mass curves on a frequency basis. The <br />latter two of these cone epts have been applied to Kansas records. Low-flow <br />frequency curves, extended to a long-term period 1920- 56, have been used to <br />compute non-sequential mass curves of natural runoff of selected frequencies <br />from which outflow is computed for various storage rates on a calculated risk <br />basis. <br /> <br />The low-flow frequency curves in figure 2 provide data for constructing <br />a mass curve of runoff on a frequency basis for the Delaware River. Imagine <br />